Closed-loop decision-making framework for electric vehicle battery recycling: Synchronizing reverse logistics network optimization with disassembly line design

The widespread adoption of electric vehicles (EVs) has resulted in a growing wave of retired power batteries, making the development of efficient reverse logistics networks (RLNs) essential for sustainable resource management and environmental protection. Considering the impact of recycling volumes on RLN, this study presents an integrated prediction-optimization framework to address the critical challenges in waste battery recycling. Specifically, the battery retirement volume prediction model is constructed and a machine learning-based decomposition-integration method is designed to achieve EV sales forecasting. Subsequently, a mixed-integer nonlinear programming model is formulated to jointly address the recycling network design and the configuration of the disassembly line, two aspects that have often been discussed separately in prior research ignoring their inherent interconnections. The formulated model incorporates a multi-operator workstation mechanism to better reflect the modular characteristics of waste batteries, with the goal of achieving more coordinated system optimization. Besides, an improved multi-stage adaptive large neighborhood search (MS-ALNS) algorithm is designed to solve the integrated optimization model. Finally, a practical case is performed to verify the effectiveness of the formulated model and the proposed decision-making framework by comparing with commercial solver Gurobi as well as the ALNS, GA, and HGA algorithms.